Universal Access To All Knowledge
Home Donate | Store | Blog | FAQ | Jobs | Volunteer Positions | Contact | Bios | Forums | Projects | Terms, Privacy, & Copyright
Search: Advanced Search
Anonymous User (login or join us)
Upload

Reply to this post | Go Back
View Post [edit]

Poster: Karl Eichwalder Date: Mar 11, 2005 11:11pm
Forum: texts Subject: Reading Error Reports on Texts

1/ Where can I read the error reports on Texts that are already posted?

2/ If I post an error report, I'm told a confirmation mail will be send to me. This does not seem to work. Does this mean, my reports will be discarded? Today I filed a report on unsershakes04eichuoft and EmilDevrient.

Karl

Reply to this post
Reply [edit]

Poster: Administrator, Curator, or Staffmolly Date: Mar 16, 2005 1:48am
Forum: texts Subject: Re: Reading Error Reports on Texts

1. We just swapped to a new error reporting system. We are in the process of creating a new system for adding error reports to the metadata.

2. Did you get an email on these? If the new system is working correctly, you should have!

-best
Molly

Reply to this post
Reply [edit]

Poster: papeters Date: Apr 7, 2005 6:58am
Forum: texts Subject: Re: Reading Error Reports on Texts

How do I look at old reports until then? Can you post the url to the old list for use in the meantime?

Reply to this post
Reply [edit]

Poster: pt Date: Apr 7, 2005 8:13am
Forum: texts Subject: Re: Reading Error Reports on Texts

We've decomissioned the old system because it became too difficult to maintain and did not provide the functionality we required. The new system is must better in every respect for us, though it does not allow you (currently) to view error reports for an item. We do not, at this time, plan on adding this functionality. I apologize if this inconvienences you.

I will look into the double slash problem. Thanks for reporting that error, though in the future please use our error reporting system. Just kidding :).

Reply to this post
Reply [edit]

Poster: papeters Date: Apr 8, 2005 3:30am
Forum: texts Subject: Re: Reading Error Reports on Texts

"...it does not allow you (currently) to view error reports for an item. We do not, at this time, plan on adding this functionality."

Well, it's certainly your call, though that makes it less likely I'd report the kind of errors I have in the past ... missing or unreadable pages in MBP texts. These are useful *if* they can be reviewed by the submitter & others. They are not useful if the secret masters want to keep the data to themselves.

As an example of why this hurts you as well as me: The reason I was looking for error reports this week is that I got the half dozen scans that a book was missing and was going to notify & offer them to you all. Can't tell if I put the error report in last year when I first detected the problems; can't review my old reports to look for the others that I've gotten fixed over the year.

Guess I'll just turn this into a suggestion that you update your PG extract so queries show both sources. If someone finds a "hit" in PG that has content that's missing from the MBP edition, someone at DP may be able to give you the missing images.

p.s. As reported at the 2002 meeting, more than 50% of MBP books are missing pages or have unreadable pages. This is true of the collection prior to that date and to new material added since.

Reply to this post
Reply [edit]

Poster: Administrator, Curator, or Staffmolly Date: Apr 8, 2005 5:51am
Forum: texts Subject: Re: Reading Error Reports on Texts

We are designing a new error reporting system specifically for texts that will become part of the metadata. Some of the books in our archive cannot be fixed by us, since we didn't scan the book in the first place. However, if some other scanning project happens to have scanned the page, we want to make it so that they can donate just the missing page.

In order for this to happen, we need to create a public error reporting system for texts, and rank it by book, scan, and processing error. For example, a missing page could be a physically missing in the book, a page that was in the book that didn't get scanned, or a page dropped while it was processed.

Anyhow, that's on our big ToDo list for how to make our text collection better. For now when we get a reported error in texts that we can't do anything about, we are sticking it in the description feild. We need to go through the old error reports and close them out and make them public.

Please don't feel like your efforts are going unnoticed! They are much appreciated.

-best
Molly

Reply to this post
Reply [edit]

Poster: tonyfirshman Date: May 4, 2009 11:41am
Forum: texts Subject: Re: Reading Error Reports on Texts

Did this project get off the ground?
There is a scanning error (or missing pages) on http://www.archive.org/details/historyofphilhar00fost

Between scanned pages 21 and 22, there should be the book's pages 3 and 4.

Reply to this post
Reply [edit]

Poster: papeters Date: Apr 10, 2005 3:59pm
Forum: texts Subject: Re: Reading Error Reports on Texts

p.s. I have to wonder how the creation of a new system is helped by the current state of things. I still create new error reports in the old format; I just can see the history any more. If it was truly a conversion issue, I would have expected a really different error *reporting* page.

Reply to this post
Reply [edit]

Poster: papeters Date: Apr 10, 2005 3:47pm
Forum: texts Subject: Re: Reading Error Reports on Texts

Thank you for your considerate reply.

"In order for this to happen, we need to create a public error reporting system for texts, and rank it by book, scan, and processing error. For example, a missing page could be a physically missing in the book, a page that was in the book that didn't get scanned, or a page dropped while it was processed."

I agree this is a thorough solution.

Could you please consider a prototype mechanism in the interim?

One that allows us to create error metadata records that indicate the book and whether the problem is with other metadate (title, publication date, etc.) or with a page (identify book page number or tif file number ... or "many/multiple" for books that have lots of problems). That and either a pick-list or free-form description of the problem should all be usable in your final solution and would allow us to make progress on tracking need & supplying replacement/additional scans in the meantime. The old format error reports will largely fit in this scheme, with minor manual intervention to extract page numbers from the error report's text.

I understand the long-term need to know the underlying cause for errors so processes can be improved. I hate to see nothing available while the grand solution is pending in a to-do list.

Reply to this post
Reply [edit]

Poster: Karl Eichwalder Date: Mar 16, 2005 2:35am
Forum: texts Subject: Re: Reading Error Reports on Texts

Thanks - yes, I received a notification, but the link contains 1 slash too much between "org" and "details":

http://www.archive.org//details/texts#forum

Terms of Use (10 Mar 2001)